Generally, network nodes or network elements such as radio base stations (RBSs), radio network control nodes (RNCs) or core network nodes, regularly conduct measurements of entities indicative of the performance in the network and collect these entities and related statistics by means of performance monitoring counters. As an example, a counter can be used to collect data representing the number of dropped calls; another counter can be used to collect data representing a number of rejected calls; another counter can be used to collect data representing a number of performed handovers; another counter can be used to collect data representing statistics on the radio resource control (RRC) success rate; and a further counter can be used to collect data representing statistics on the throughput (uplink or downlink throughput) in a cell. The data from each counter is usually stored separately in a database or a register either as the average over a certain time or as a probability density function, based on the periodic sampling of the measured data (or entity). The stored data can be subsequently analysed by a network operator to verify the configuration of the network. The number of databases may also increase with the number of counters. Furthermore, the databases have little or no communication between them. This makes it hard for the operator to manage the network performance. For example, network operators would currently have difficulty associating a sudden drop of the RRC success rate with a change in uplink or downlink throughput that caused it. This also means that it is currently difficult to adequately reconfigure or adapt network characteristics to e.g. balance the load in the network or to dimension network resources etc.
Consequently, there is a need for an improved solution which simplifies for the network operator to associate a plurality of measured entities indicative of the performance of the network in order for the operator to adapt/optimize it network.